Image Sharpness Estimation
==========================

The algorithm performs image sharpness estimation on a set of 2D images.

Input
-----

* Single image or a set of 2D images.

Output
------

* Blur annoyance coefficient ranging between 0 and 1 for each input image, where 0 indicates the sharpest quality (least blur annoyance) and 1 indicates the worst quality (highest blur annoyance).

Description
-----------

The method measures blur annoyance coefficient in an image by comparing pixel intensity variations before and after applying a low-pass filter. Large differences indicate a sharp original image, while small differences suggest the image was already blurred. This is based on the premise that blurring has a stronger impact on sharp images compared to already blurred ones.

The algorithm has one parameter:

* **Blur kernel half size**. Defines half the width of the blur filter kernel, which determines how many neighboring pixels are considered when applying the blur.